(navigation image)
Home American Libraries | Canadian Libraries | Universal Library | Community Texts | Project Gutenberg | Children's Library | Biodiversity Heritage Library | Additional Collections
Search: Advanced Search
Anonymous User (login or join us)
Upload
See other formats

Full text of "Merging information technology and cultures at Compaq-Digital : case study"

MIT LIBRARIES DUPL 

iiiiiiiiui mil mil lilll nil 1 1I1IIII 



3 9080 02246 1815 



328 \U i OEVi 

414 "^ ^ - 

Mi iT- 



i/i; 



CASE STUDY 

Merging Information Technology and Cultures 
at Compaq-Digital 

Jeanne W. Ross 
Carol Brown 

January 2001 

CISRWPNo.312 

Sloan WP No. 4155 i^ 



©2001 Massachusetts Institute of Technology. All rights reserved. 



Center for Information Systems Research 

Sloan School of Management 

Massachusetts Institute of Technology 

3 Cambridge Center, NE20-336 

Cambridge, MA 02142 



MASSACHUSEHS INSTITUTE 
OF TECHNOLOGY 



APR 1 8 2001 



LIBRARIES 



CISR Working Paper No. 312 



Title: Merging Information Technology and Cultures at Compaq-Digital 

Author: Jeanne W. Ross and Carol Brown 

Date: January 2001 

Abstract: Compaq Computer Corporation and Digital Equipment Corporation formally 
joined forces on June 11, 1998, creating a $37 billion (US) personal computer and computing 
services firm. The merger was expected to generate synergies that would lead to cost take-out 
and cross-selling opportunities. In order to generate these benefits, however, management of the 
combined firm would have to overcome the challenges presented by two very different 
management styles, cultures, organizational forms, and technology infi-astructures. This case 
describes the challenges from the perspective of the two firms' IT units. It explains how IT 
leaders ft-om the two firms cooperated to deliver an ambitious set of Day One goals, and how 
they worked through ongoing challenges during the eight months immediately following the 
merger. It offers lessons about the role of technology in facilitating merger-induced 
organizational changes and serves as a reminder of the importance of strategic direction in 
setting the IT agenda. 

22 Pages 



This paper is also available as Sloan Working Paper No. 4155. 




Center for Information Systems Research 
Sloan School of Management 
Massachusetts Institute of Technology 



Merging Information Technology and Cultures at Compaq-Digital (A): 
Meeting Day One Objectives 



On June 11, 1998, stockholders of Digital Equipment Corporation approved its acquisition by 
Compaq Computer Corporation, creating a $37 billion (US) personal computer and computing 
services firm. Within an hour of the stockholder vote, the more than 100,000 employees of the 
new Compaq were able to access a smgle e-mail directory via a corporate network that linked its 
world-wide sites. The IT units of the individual firms had banded together to successfully deliver 
their Day One objectives: 

There wasn't a whole lot of time to decide who was going to make what decision. 
As a team we marched off to figure out how we were going to get the networks 
to talk together even though they weren't the same technology. We figured out 
how to share mail, messaging — that kind of thing. It just had to happen pretty 
quickly. So we had a common task that we both went after. 

— Fred Jones. Vice President, Information Services. Compaq 

Experts inside and outside the firm recognized that the Compaq-Digital merger created a firm that 
could rival Hewlett-Packard and even IBM. But in order to leverage the potential synergies, 
Compaq's management team would need to overcome the challenges posed by the two firms' 
very different business models and cultures. 



Company Backgrounds 

Established in 1957, Digital Equipment Corporation introduced minicomputers to corporate 
computing environments. As minicomputers encouraged greater distribution of computer 
processing in organizations. Digital rose to become the world's second-largest computer maker in 
the mid-1980s. Headquartered in Maynard, Massachusetts, USA, Digital flourished under a 
highly distributed structure that encouraged innovativeness and consensus decision-making. In 
1 997 the company had over 1 1 00 sales, production, and service sites world-wide, of which 400 
were in the U.S. 

In the 1 990s, management underestimated the importance of microcomputers in business, and the 
firm sank to a weak fourth among US computer manufacturers. Management responded by 
shifting business focus from computer manufacturing to computing services. In 1997 Digital had 
revenues of $13 billion, almost half of which were generated by its services business. More than 
22,000 of its 70,000 employees were service business employees. At the time of the merger the 
firm was the third largest provider of corporate computer services after IBM and Electronic Data 
Systems (EDS). In 1998, as it was about to merge with Compaq, Digital's profits exceeded $1.3 
billion. 

This case was prepared by Jeanne Ross, Principal Research Scientist at MIT's Center for Information 
Systems Research, and Carol Brown, Associate Professor at Indiana University. The authors would like to 
thank Rick Fricchione of Compaq for his insights and efforts and the many Compaq employees who 
participated in the research. 

©2001 Massachusetts Institute of Technology 



Compaq Computer Corporation was created in 1982 to clone and sell personal computers. By 1997 the 
company had three primary Imes of business: (1) consumer PCs, (2) commercial PCs, and (3) enterprise 
PC servers. Compaq was highly centralized with almost half of its 40,000 employees located at its 
wooded campus in Houston, Texas, USA. The company had been growing through geographic expansion 
and by the end of 1997, 40% of its $25 billion in revenues were generated outside North America. 

Compaq's competitive niche was as a high-volume, low-cost manufacturer. Compaq had no services 
organization; its primary business model was to ship fairiy standard products in large volumes to channel 
partners who configured them for their end-customers, hi North America, for example, 90% of its sales 
were with 35 partners. Its chief competitor, Dell Corporation, was moving the industry to a build-to-order 
model. Consequently, one of Compaq's key challenges was to transform its business model so that it 
could build and ship computers according to customer specifications for an order placed via the Web. 

Compaq's CEO Eckhard Pfeiffer, announced early in 1997 that he wanted to double its sales revenues to 
$50 billion by 2000. Towards that end, Compaq first acquired two small technology firms. Then in mid- 
1997, Compaq announced its intentions to purchase Tandem Computer Corporation, a $2 billion 
manufacturer of large, customized computer systems. Wildly successful in the eighties. Tandem had 
fallen on hard times. By the end of 1997, the Tandem deal was complete, and Compaq turned its attention 
to an even bigger acquisition. 



Announcement of the Merger 

Compaq announced its intention to buy Digital on January 26, 1998. To most employees in the two firms, 
the news came as a surprise: 

The date the official announcement was made it just floored us all. If you have been in 
this industry for a long time, like I have, IBM and Digital are the icons of the computer 
industry — forever and ever. This was a company that is a legend in our industry. This is 
Ken Olsen. This is the big red building in Maynard. 

— Ed Pennington, Director, Corporate Administrative Systems 

Compaq management explained to investors that the benefits of the acquisition stemmed largely from the 
synergies of merging Compaq's personal computer and server manufacturing business with Digital's 
computing services busmess. Specifically, it would lead to (1) cost take-out by eliminating redundancies 
in computer manufacturing capabilities and (2) cross-selling opportunities through the merger of the 
firms' customer bases worldwide. The deal was to be accretive within one year, meaning that Compaq 
would have incurred all merger-related costs and the benefits would start to accrue. 

It was clear to insiders and outsiders alike that this would not be an easy merger. Although the two firms 
were in the same industry, in practice they had very little in common. For example, Compaq 
manufactured to replenish inventory; Digital manufactured to order. Compaq had a single level Bill of 
Materials; Digital had a ten-level nested Bill of Materials. Compaq was organized around geography; 
Digital was organized globally around line of business. Compaq was a warranty business using partners 
to take care of the customer; Digital was a contract business with a large internal customer service staff. 
Digital's high end server and PC business was largely an enterprise account focus. 

In addition, the two companies had very different management styles. Compaq was well-known for its 
rapid decision-making culture. Management would make a decision, stick to it, and go with it "life or 
death." Outsiders referred to it as a Texas cowboy culture. Digital had a consensus approach to 



management. Decisions were often made by committee. Rick Fricchione, a veteran Digital IS manager, 
described the difference: 

Compaq rewarded quick decisions and heroics to make them happen. Sometimes this 
worked just fine. Other times the chewing gum in the hole dried out and you found 
yourself back in the same position (or worse). Digital tried to find systemic solutions. 
When we shot something, it died. It was thorough, but much slower. Compaq was 
probably more successful in hitting the market window for something, but at a cost of 
very fragile business systems. 

— Rick Fricchione, Vice President, 
Advanced Information Technology and Planning 



Pre-Merger IM Integration Efforts 

Although the initial announcement of the acquisition suggested that Compaq might manage Digital as a 
separate strategic business unit, the Board soon articulated its intention to create a new integrated 
computer company that would deliver enterprise-wide FF solutions with the efficiency of a PC company. 
Both firms established an Integration Office that reported to their respective Executive Committees. The 
two firms' Integration Offices had parallel Integration Teams for each of their key functions: 
finance/administration, services, human resources, manufacturing, sales & marketing, and information 
management. 

Executives of the Information Management (IM) organizations at the two firms recognized that IM would 
play a critical role in the transformation from two independent companies into a single, integrated firm. 
In addition to the two CIOs, the IM integration teams at both firms were led by two ClO-reports with 
broad responsibility for IT infrastructure: Mike Baker and Fred Jones at Compaq, Tom Simmons and 
Rick Fricchione at Digital. (See Exhibit A- 1 for the Integration Team structure.) 

The IM integration team worked towards integration of the two firms' IM units. This was in marked 
contrast to the integration plan for the earlier Tandem acquisition. Tandem was operating as a subsidiary 
and Compaq's CIO had issued four directives to Tandem IM management: (1) convert Tandem's 
proprietary email system to Microsoft Exchange and integrate with Compaq's email system; (2) convert 
to Compaq's telecommunications system; (3) abandon Oracle financial systems and implement SAP; and 
(4) implement the same manufacturing systems as Compaq. Tandem IM managers were frustrated by the 
initial lack of integration and were pleased when eventually the Tandem unit was also merged into the 
new Compaq organization: 

Nothing was getting replaced; nothing was getting merged because there was too much 
independent thought going on. So having the groups become one — all of the financial 
programmer analysts and management from the Tandem side getting merged in and 
working directly for the same fimctional IM group in Houston — finally let us stop 
continually modifying those separate solutions and start toward one solution. 

— CIO, Tandem 

The immediate integration of the Compaq and Digital units averted similar frustrations, but it involved 
combining units with very different structures and operating styles. Digital had a federated FT 



' Compaq's IT unit was called Information Management. Digital's IT unit was called Information Systems. 

Fred Jones, who headed up infrastructure responsibilities for Compaq, had the title Vice President for Information 

Services. 



organization structure in which business units owned their applications but infrastructure was under the 
control of CIO Dick Fishbum. Prior to Fishbum's arrival as Digital's first CIO, IT had been the purview 
of the individual business units. Fishbum created a single IT organization, but left the money in the 
business units so that the corporate IT budget was defined by business unit needs. He introduced shared 
services into Digital IT operations, which reduced IT costs from 6°/o of revenues to about 4%. He was 
comfortable in Digital's consensus culture, which, according to one Digital-Classic^ IM manager, treated 
a committee vote of 17 to 1 like a tie: 

Dick was a cerebral type who rarely raised his voice. When he spoke, he usually had 
something to say. In most cases, Dick would convince you of his position. Only once in 3 
years did I see him reverse someone's decision. 

— Rick Fricchione, Vice President, 
Advanced Information Technology and Planning 

In contrast, Compaq CIO John White ran a highly centralized, top-down organization in which he was 
known to make decisions several levels down into the organization. Computer operations and global 
telecommunications management were Houston-centric. A highly integrated infrastructure ran Compaq's 
high volume, build-to-stock business, and individual business units had no control over their IT costs. 
Under White, the IM organization had a 1997 budget less than 2% of sales.^ White had a preference for 
speed, which fit well with Compaq's reputation for throwing 100 balls into the air, catching about 80 of 
them, and then cleaning up on-the-fly. 

High-level conversations between senior IM executives at the two firms started soon after the 
announcement of the acquisition. In early February 1998, four of Digital's senior IM managers went to 
Compaq's headquarters to lay the groundwork for the eventual integration of their two organizations. Due 
to legal restnctions on the kinds of information that could be shared, the discussions about existing 
systems and practices were very high level. In essence, each side shared presentations that they regularly 
made to customers. The meeting primarily served to help familiarize the two parties: 

1 took notes but was told 1 couldn't share them. So when 1 came back, a lot of what I 
shared was "tone" — like what it was like to walk onto their campus. It felt like they were 
moving 800 miles an hour. They weren't fully staffed a lot of the time, and they were 
looking for ways to get things done. 

— Technology Manager, Digital 

At a follow-up meeting in late February, a dozen or so of John White's direct reports met with Dick 
Fishbum and a number of his direct and dotted-line reports. The groups identified 15 key areas that 
needed to be integrated. They paired counterparts from the two firms' IM functions in these 1 5 areas to 
start working on an integration roadmap for both the infrastmcture and mission critical applications. 

They then mapped out 44 integration initiatives and created teams to work on them. (See Exhibit A-2 for 
a list of the key areas and initiatives.) Although the focus of the meeting — and the teams — was on 
technical integration, issues around organizational stmcture quickly surfaced: 

The first thing that came to everyone's mind was that there was a huge disparity in 
staffing. Digital was gigantic. John [White] basically said, "You have hundreds of people 



^ The terms Compaq Classic and Digital Classic were used to signify the pre-merger entities. 

^ The percentage of revenue figures for Compaq and Digital are not totally comparable due to different accounting 

practices around software and project cost depreciation. 



too many in this group." I think Compaq was just overwhelmed by the shear size of the 
Digital groups/ 

—CIO. Tandem 

John White charged the teams with evaluating Compaq's and Digital's systems and processes. They were 
to present recommendations m May for adopting the Compaq, Digital, or hybrid solution for each major 
initiative. IM Integration leaders instituted a weekly review process for ensuring that the teams were on 
track, and the teams worked cooperatively towards their goals. Although Compaq was the acquiring firm. 
Digital had more human resources to devote to the planning. 



Developing an Integrated Infrastructure 

Among the 44 IM integration initiatives, management was particularly concerned with those that would 
establish a common infi^structure. Even before the merger was formalized, the two firms needed an 
efficient, secure communications channel for exchanging information between them. And soon after the 
merger, management wanted a technical infrastructure that would allow it to operate as a single company. 
To this end the IM Integration leaders assigned teams to nine initiatives that they referred to as Phase One 
— actions that were to be implemented within sixty days after the merger was formally approved. Another 
16 initiatives were referred to as Phase Two — planning efforts for post-merger activity. Enterprise 
Network Integration, Enterprise Application Architecture, and Day One Deployment were examples of 
Phase One initiatives. 

Enterprise Network Integration 

Laurence Cranwell, manager of global communications for Digital, headed a team responsible for 
planning the enterprise network for the combined company. Cranwell's team had the challenge of 
ensunng that a packet could reach across any part of the existing Digital or Compaq (including Tandem) 
networks, even if they had different addresses and different architectures. The two companies had 
different network architectures, products, and protocols. Compaq's network architecture had been 
developed for a highly centralized operating environment, whereas Digital's Services business required 
more distributed operations: 

We [Digital] had 400 sites in the U.S. They had ten. That's one thing we worked through 
as an organization. The stove pipe organization that does technology delivery is okay 
when you've got very few sites, but when you have many, you need an operating model 
for 1 100 global sites. You can't be there all the time. 

— Laurence Cranwell, 
Director, Communication Technology Information Services 

Digital's networking people viewed their responsibility for the global network as an opportimity to do a 
"green field" design for the future. Both firms' legacy networks had become congested and both were 
working with earners to design ATM networks, so these efforts were combined. They would need six 
months to turn on the green field, but they needed to design the global network in advance of the actual 
merger. This involved making assumptions about future site locations: 

A lot of the future communications infrastructure would depend on the data center 
decisions — where they are going to be — and where the transactions are going to take 



^ Digital had about 3200 people in its IT organization the day of the merger. Compaq had about 1500 IM en^jloyees 
plus 500 contractors. Tandem had a few hundred. 



place. Plans also depended on what applications would be selected to go forward, and 
which plant sites would remain open or closed. Plant closings were very sensitive issues; 
management didn't want the information to get out and scare people off prematurely, but 
we needed to know. 

— Laurence Cranwell 
Director, Communication Technology Information Services 



Although the uncertainties created challenges for planning the network, the network team could outline a 
high-level architecture based on the Board's statement that Digital would be integrated with Compaq. 
They needed to know less about the structure and operations of the combined firm than the application 
integration teams would need to know: 

It's really like train tracks. They conform to standards, but the types of trains you can run 
over the tracks are pretty different. That's where you get a little more disagreement. You 
don't get into the politics as much about the tracks. You take care of the tracks — make 
sure I can get from point A to point B — and others worry about which trains, the 
application level. 

— Ed Furilla, Sales & Marketing IM Manager 

While they planned the new global architecture, Cranwell's team worked to establish temporary linkages 
across the three entities. Compaq had nine CLASS B internet addresses, and about 80 CLASS C internet 
addresses.^ Digital had one large flat address space. Consequently, if Compaq, Digital, and Tandem all 
had buildings in Munich, consolidating their facilities could result in having 3 people sitting twenty feet 
apart with 3 different IP address domains supporting them. Until the firm had a new architecture with a 
single address space, communications between the merged entities had to be transferred through internet 
mail (i.e. through firewalls) rather than a seamless network. This was both slow and costly: 

The joke was that Compaq paid $9.6B for a CLASS A internet address. The right answer 
was to redesign the network to move all of Compaq into the CLASS A address that 
Digital had, and at the same time re-architect. 

— Rick Fricchione, Vice President, 
Advanced Information Technology and Planning 

Enterprise Application Architecture 

Because Compaq was the acquiring firm, its application systems were the starting point for discussions 
about which of the two firms' redundant systems they should retain. At the time the merger was 
announced, both firms were in the third year of SAP R/3 implementations to replace core transaction 
processing systems for finance, manufacturing, order management, and logistics. Both firms had focused 
first on European operations because they believed financial returns would be most immediate there. But 
the two firms had little else in common in their ERP implementations. 

Digital was implementing a common, global supply chain system across its business units. It was 
pursuing a site-by-site implementation managed out of a central program office that leveraged the 
expertise of about 100 key people across the organization. A key objective of the SAP implementation 
was to develop consistent business processes and consistent reporting across sites. Digital had allowed 



' The CLASS refers to the size of the address space. A CLASS C address space might allow for as few as 256 nodes 
on a network, while a CLASS B address could allow for 64,000 nodes. Large firms must cobble together their 
networks using multiple address spaces if they do not have a sufficiently large address space. 



just 14 modifications to SAP code. Digital had implemented SAP financials worldwide, standardizing all 
financial processing such that the way revenue was counted in Asia was the same as in Europe. Following 
about three false starts, Digital implemented SAP in its European manufacturing operations, and by June 
1998 Digital had installed four instances of SAP manufacturing modules, including one US 
manufacturing site. 

Compaq's implementation was managed out of the IM organization. The cenft-al team developed a 
template and each geographic region customized the template to fit local needs. By June 1998, the three- 
year project had included about 40 small "big bang" implementations with up to 10 projects going on in 
parallel. In total, it had nine separate instances of SAP. Compaq had relied heavily on Price Waterhouse 
consultants and had focused on a fast implementation rather than consistency across sites. Compaq had 
made over 1500 modifications to SAP code as it customized the software to meet the needs of individual 
sites, plants, and geographies: 

We put stuff in brutally fast. We were almost up 100% on SAP before we bought Digital. 
We would just go in and make people scream; if the users didn't like it, we told them 
'you're going to get it anyway.' Digital was just the opposite: it was very much risk 
avoidance. They wouldn't go live unless all the issues were resolved, and everybody 
signed up that it was going to work. When they went live, there was never an issue. Then, 
again, they had very little SAP up by the time we bought them. 

— Director, Global Financial Processes 

Having a common software package did not make for an easy integration of the two firms' operating 
processes. For a variety of reasons, including the fact that Compaq was the acquiring firm and felt that it 
was further along in its implementations, the SAP team eventually recommended that the firm pattern 
future implementations after Compaq Classic. A short-term dilemma that arose following the SAP 
decision was whether to continue with a Digital SAP implementation scheduled for Summer 1998 in 
Scotland. Despite the stated new direction, the uncertainty of the timing of the merger led to the decision 
to implement in the Scotland plant as planned. 

As part of its SAP implementation, Compaq had completed a two-year rollout of the HR module in 
March 1998. In conti-ast. Digital, along with Tandem, had purchased PeopleSoft for its HR systems. Both 
SAP and PeopleSoft were asked to propose new contracts for the combined firm. Although tempted to 
push SAP into all of Compaq, the IM and human resource organizations chose to adopt PeopleSoft. This 
was the first time that Compaq employees had been in a situation where an alliance affected a platform 
choice: 

PeopleSoft is a business partner. We were looking for an opportunity to expand our 
relationship with PeopleSoft so that became part of the decision critena. It was also 
important that, once the acquisition took place, 70% of our workforce would already be 
on PeopleSoft. 

— Ed Pennington, Director, Corporate Administrative Services 

At the time of the PeopleSoft decision. Digital was 80% through its HR implementation. However, the 
decision was made to scrap this implementation, because it was not consistent with Compaq's processes. 
The combined firm would start all over again with a PeopleSoft implementation for the new Compaq. 

The differences in Compaq's and Digital's approaches to enterprise application implementation were 
echoed in the differences in the two firms' approaches to enterprise application support. Digital had 
assigned responsibility for supporting its PeopleSoft implementations — and most of its other global 
legacy systems — to a large shared services organization called the Production Support Group (PSG). At 



the rime of the merger the group totaled about 1000 people, and funcrioned like an in-house outsourcing 
organization. Digital credited the PSG concept with reducing IT costs and moving spending from 
operations support to new investment because it forced business unit managers to decide how much 
support they were willing to pay for. 

Compaq located responsibility for global application management within each of the geographies. Some 
key managers were philosophically opposed to the shared services concept, believing that it introduced 
unnecessary bureaucracy. Thus, PSG had no counterpart at Compaq and its head was not paired with a 
Compaq manager for integration planning. The design for legacy systems support defaulted to the 
Compaq model: 

The key issue is not shared services, but running the corporation in a geo model or a 
business unit/functional model. Compaq had chosen geography. Once that decision is 
made, other things come as a result. 

— Rick Fricchione, Vice President, 
Advanced Information Technology and Planning 

Day One Deployment 

IM managers at the two fums worked closely to ensure some level of IT integrarion of the two firms as of 
the day the merger would be approved by the stockholders. Leveraging their experience with the Tandem 
merger, Compaq executives specified some aggressive Day One goals: 

One of the things done enormously well for the Tandem acquisition was to get the two 
networks talking within two days. So our goal with Digital was to have this capability on 
the day the merger was approved. In some of the other systems areas, the need to take 
immediate action wasn't quite as important. We were able to think about it a little longer. 

— Fred Jones, Vice President, Information Services, Compaq 

In particular, the firms took advantage of the fact that they were both using MS Exchange for email. On 
Day One a single email directory was available so that employees could easily find and mail messages to 
their counterparts in the other organizarion. Firewalls and filters that had been put in place prior to the 
merger came down to provide open connectivity, with no security holes to the outside world. A 
coordinated approach to help desk support combined Compaq's help center in Houston and Digital's in 
New England. And a new external web site put a combined front-end to the two firms" individual sites. 



First Day as New Company 

The two IS organizations viewed having the new email infi^structure in place on Day One as a 
collaborative triumph. The integration teamwork had also increased sensitivities to the cultural 
differences across the two organizations: 

You would hear all kinds of stories like: 'Why do those folks in New England think they 
have to plan forever? We could have it done before they have a plan.' I heard comments 
like these in the first two months, but 1 didn't hear them thereafter. You could almost 
make these deficiencies something to be proud about on both sides, because if you mixed 
the two, you'd have planning and execution. 

—John Buda, Director, IS Client Management 



In addition, the pre-merger work resulted in some important planning decisions that would guide systems 
integration m the coming months. However, the Digital IT people who had worked on the integration 
teams did not know whether they would be tapped to play a major role in the new IM organization or not. 
Throughout the planning, the fact that leadership was shared by paired teams of IT managers meant that 
the temporary organization chart effectively had a "two in a box" look that would soon be reduced to one: 

There was plenty of tension, but it wasn't as intense as I would have expected it to be. 
We worked well together, especially the Information Services group. I think it was 
because there was compelling work that had to be done. The Day One set of tasks had to 
happen on schedule. 

— Fred Jones, Vice President, Information Services, Compaq 

The public documents filed for the merger made it clear that 16,000 jobs would be cut from the new 
Compaq by the end of 1998, and anxieties were high, especially among those people who had been with 
the company the longest. The fiscal year for Digital Classic would end June 30 and, due to its strong 
financial results, the year-end bonus looked attractive. However, to receive the bonus, employees needed 
to stay with the company until the end of July. 



Exhibit A-1 



Merger Roles/Responsibilities 



Ilene Jacobs - Digital HR 

Digital Integration Teams 
- Finance/ Admin 

-IM 

- Services 
-HR 

- Manufactunng 

- Sales & Marketing 



Rick Fricchione - Digital 
Tom Simmons - Digital 
Fred Jones - Compaq 
Mike Baker - Compaq 

IM Technology & Operations 
Integration Teams 

-44 task teams 



Compa()/Digital Executive Cranmittee 



Mike Baker - Compaq 
Tom Weyant - Digital 



Integration Coordinator 



Dick Fishbum - Digital QO 
John White - COMPAQ CIO 



Hans Gutsch - COMPAQ HR 

Compaq Integration Teams 

- Finance/ Admin 
-IM 

-Services 
-HR 

- Manufacturing 

- Sales & Marketing 



Digital/Compaq IM StatT Members 
With AIT? Support 

IM Application Integration Teams 

(by aligned IM organization) 



Digital/Compaq Integration Coordinator 
IM Staff Members plus selected others 

Digital Friday Integration Status Meetings 
Compaq Monday 



Exhibit A-2 



IS/IM Integration Teams - Initiatives 



1 . Information Management Vision 

2. Human Resources IS Workforce 

3. IS Management Team Exchange 

4. SAP R/3 Centnc Business 

5. Applications Portfolio 

6. Intranet/Internet Deployment 

7. DIGITAL Product and Services 
Offerings 

8. Infrastructure Topology 

9. Start-up Deployment Team 

10. Mail and Messaging Integration 

1 1 . NT Infrastructure Integration 

12. WAN Optimization 

13. Remote Access 

14. LAN Strategy 

15. Enterpnse Network Architecture 

16. Enterprise Management 
Architecture 



•17. Voice Network 

•18 Internal Equipment Sourcing 

•19 Disaster Recovery Strategies 

•20. IT Metncs 

•21. Technology Integration Checklist 

•22 Data Warehousing/Reporting 

•23. Services Logistics 

•24. Call Management 

•25. Enterprise Architecture for SAP 

•26. Peoplesoft/SAP HR 

•27. Year 2000 Program 

•28. Project Services 

•29. Reference Data 

Strategy/Management 
•30 Sales Productivity Tools 
•31 High Availability/High 

Performance Computing 



•32. Operational Processes 
•33. Information Security 
•34 Desktop Support Strategies 
•35. Collaboration and Workflow 

•36. Cost Recovery for IS and 

Infrastructure Services 
•37. Merged SAP Enterprise 

Application Architecture 

•38 Alliances and Agreements 
•39. Sales Workbench and Transfer 

Manager 
•40. Product Management and 

Development Workbench 
•41 Document Management and 

Records Retention 
•42 Electronic Document Interchange 
•43. Treasury Management 
•44. Worldwide Trade and Compliance 



Merging Information Technology and Cultures at Compaq-Digital (B): 

Becoming a Single Firm 



In August 1998, within two months of the merger of Compaq Computer Corporation and Digital 
Equipment Corporation, a new Chief Information Officer, Michael Capellas, was brought on board. 
Capellas arrived with a track record as an international CIO and recent stints as Director of Supply Chain 
Management for SAP America and a senior VP within Oracle Corporation. At that time, he was the only 
senior manager within the newly integrated computer company that had been brought in from the outside. 

When Capellas arrived, the basic communications infrastructure, telephone systems, and e-mail were in 
place, but integration of the daily operations — taking orders, shipping orders, interfacing with 
customers — was still underway. Nonetheless, within four months after his arrival, Capellas had seen the 
new company set a world record for the number of units shipped — 4.2 million; 

Despite all the problems, we are outperforming the world in volume. At the end of the 
day, business results are business results. We must be doing something right. 

— Michael Capellas, Senior Vice President & CIO^ 



Creating a New Information Management Organization 

The new Compaq was a combination of the old Compaq's high-growth personal computer and server 
businesses with Digital's mid-range computer products and highly profitable services business. 
According to documents filed with the U.S. Securities and Exchange Commission as part of the merger 
approval process, the new company was expected to achieve lower operating costs as well as increased 
revenues from new cross-selling opportunities. Senior management felt pressure to deliver these benefits 
quickly, and the Information Management (IM) organization's new leader felt accountable: 

In our business IT is not an enabler, it basically is the business. If you're not getting 
orders, you're not shipping product... and the world is ending. 

— Michael Capellas 

Under its former CIO John White, Compaq had had a highly centralized approach to managing IM. This 
structure helped the IM organization get things done quickly and kept IT spending significantly below the 
industry average. Although each of the geographic regions had their own IM support groups, White was 
tightly linked with Compaq's executive committee and IM strategy and infrastructure standards decisions 
were all set at the corporate level. 

In contrast, under CIO Dick Fishbum, Digital had had a highly decentralized model, with each business 
having its own CIO. This model gave the distributed manufacturing and services businesses more 
personal service from their own IS groups, but it was more expensive. It also meant that the business 
CIOs often went off and did something on their own. 

The June 1998 post-merger announcement of the top management team for the new Compaq named 
White as CIO — but it was well known that White planned to retire sometime before year-end. In fact, the 
merger integration activities had thrown a 'kink' in the search process for a new CIO and White had 
signed an extension to his contract. The agreement was that he would stay on until a replacement was 



' The interviews for this case study are based on on-site interviews between December 1998 and February 1999. 



11 



found and the merger activities had settled down a bit. This left plans for the new IM organization 'in 
limbo' over the summer months. 

Everyone was concerned about 'where do I fit on the org chart.' That's what they really 
wanted to hear. So until we could actually tell them that, we gave them a lot about how 
the business is commg together, what other organizations at Compaq were doing, and 
where we were on our plans. 

— Laurence Cranwell, Director, 
Communication Technology Information Services 

TTie new senior IM management team was named by White just one week before he retired in August 
1998. Thus, when Capellas arrived from Oracle Corporation to head up the combined IM organization, he 
inherited a new management team that had not yet sorted out lower level reporting structures. 

White chose primarily Compaq Classic managers for his senior management team, but as Capellas filled 
in the organization chart (see Exhibit B-1), it became clear that IM management at Compaq would no 
longer be so tightly controlled from Houston. Rick Fricchione, for example, was an ex-Digital manager 
who was tapped to be Vice President, Advanced Information Technology and Planning, and Capellas had 
agreed that Fncchione could fill the role from Massachusetts. Similarly, Fred Jones, a Compaq Classic 
manager who assumed global responsibilities for technology and process, had five former Digital 
managers among his eight direct reports. Only two of the five had plans to relocate to Houston. Laurence 
Cranwell, one of Jones' direct reports based in New England, managed a team split about 50/50 between 
Compaq Classic and Digital Classic. 

While Compaq Classic managers lamented the increased complexity that accompanied a more distributed 
IM management model, they acknowledged that, in a very tight IT labor market, Houston alone could not 
provide sufficient FT resources for a $40 billion firm. They wanted to take advantage of the fact that 
Compaq's two key acquisitions were in high tech corridors. The 1997 Tandem acquisition gave Compaq a 
hook into the Silicon Valley, and the 1998 Digital merger gave it a hook into the Boston comdor and 
Route 128. 

Adding to the complexity was the recognition that Compaq's newly acquired Services business had 
different IM support needs than the manufacturing business. CEO Pfeiffer had kept the leadership for the 
Services business in Massachusetts, reporting to both the CEO and the geographies. The Senior VP of 
Services John Rando had been a key player in the strategy discussions during the pre-merger integration 
process, and Rando became a member of the new Compaq's top management team as soon as the merger 
was fmalized. The IM roadmap for Services mirrored this structure. Mike Whitney, the former CIO for 
the Digital Services group, reported to Rando, and the IM teams working on worldwide and fiinctional 
systems solutions for Services reported to Whitney. The IM heads working in each of the geographical 
Services units now had a dual reporting relationship to Whitney and to their local Services head: 

On Capellas' new organization chart, I'm the only person who is shown with a dotted 
line; everybody else is a solid line. Services makes a large percentage of the profit of the 
company. But it's difficult in terms of being remote from Houston and there's still 
concern on how well Services is perceived as integrating into the rest of the company. 

—Michael Whitney, CIO, Compaq Services 

While Houston IM managers were adjusting to the challenge of managing a distributed work force, those 
former Digital managers who had decided not to relocate to Houston had to learn when they could rely on 
e-mail and video-conferencing and when they needed to be in Houston. They knew that they needed to 



12 



interact with other IM and business managers in order to build good working relationships, and they had 
to periodically be part of the action in Houston in order to influence decisions: 

Things can happen so fast, that if you're not in the middle of the floor, you're out of the 
decision process. The remote manager suffers a little from out-of-sight, out-of-mind, 
because it's been the mentality of the company. 

— Michael Capellas, Senior Vice President & CIO 



Working Toward Process Integration 

In order to achieve the objectives of the merger, senior management felt that process integration was 
critical. Immediately following the merger, a customer who wanted to buy Compaq, Tandem, and Digital 
products would have to submit three separate orders and later process three separate invoices. Compaq 
worked to provide a single customer interface within a couple months of the merger, but real process 
integration would take some time: 

The company had done a pretty solid job of merging the sales force, but the infrastructure 
was not in place to [enable all salespeople to sell all Compaq products]. The order 
fulfillment model still had a long way to go; it was just in its infancy. Finance was getting 
the books closed, but with some brute force. 

— Michael Capellas 

Ultimately, management wanted an integrated supply chain across the three entities. The business vision 
was that a customer could call Compaq and order three DEC Alpha's, two Tandem Himalayas, and 10,000 
Compaq PCs on one purchase order. The order would be reflected as one entry in Compaq's systems and 
would arrive at the customer's location all at one time. To achieve this level of integration, management 
had to (1) reconcile business model incompatibilities among the separate entities, and (2) implement 
integrated systems to support those enterprise-wide processes that defined the new Compaq. Management 
coined the term OnePaq for this integration effort. (See Exhibit B-2.) 

Reconciling Different Business Models 

Compaq's core business processes supported large manufacturing volumes — as many as 20,000 PCs in a 
day in a single plant, which was more than Digital manufactured in all its plants in a week. Digital's 
business model, on the other hand, supported direct-to-customer sales and support of smaller numbers of 
PCs (up to 3000 per day) and customized mini and mainframe systems. Compaq had acquired Digital 
specifically because of its Services business, but the opportunities it presented and the challenges of 
integrating services into supply chain, customer facing, and e-business processes were not well 
understood. 

The Services business negotiated high-end systems contracts directly with enterprise customers. Compaq 
Classic had not sold services. Its services unit had included a help desk, repair center, and spare parts 
shipping for its products. Rather than deal with end users, Compaq supported its channel partners who 
supported Compaq's customers. In contrast. Digital had serviced its own systems and that of its 
competitors. In addition, Digital had sold systems integration services, while Compaq had regularly used 
the services of systems integrators like Andersen Consulting and PricewaterhouseCoopers. These very 
different business models made for an uneasy transition following the merger. Some Compaq Classic 
managers sensed that Digital's willingness to service competitors' equipment compromised potential 
Compaq sales, while former Digital managers were wary of the impact of Compaq's alliances with 
competing service providers on Service unit revenues. 



13 



Ultimately, Compaq Service engagements that involved implementing systems like SAP, Siebel, or e- 
business applications would "pull" Compaq hardware sales. Similarly, hardware sales were expected to 
lead to Services revenues. Because of the differences in old business models and systems, however, 
Whitney anticipated that it would take some time to see synergies from the merger of the Compaq, Digital 
and Tandem Services units. The integration would involve a global portfolio mapping of all Compaq 
services across the world and the implementation of the SAP baseline to support them. 

An early initiative immediately following the merger was a Web-based front-end for on-line orders via 
the Internet. This, of course, did not clean up the back-end systems. One Vice President noted that the 
new front-end was like "putting lipstick on a pig." But the web-based front-end was important for moving 
Compaq Services into the Internet space: 

The Internet world sort of changes the spin on how you want to deliver services. For the 
small-medium business environment, more than likely we'll look at a lot of Webbing and 
a lot of the Compaq Classic stuff, because we can't afford $100-200K service centers. 
But the time that it's going to take to integrate all the call-handling, all the sales and 
marketing, is nowhere near the speed that the business wants to go. 

— Michael Whitney, CIO, Compaq Services 

The growing importance of e-business exacerbated the organizational challenges associated with the 
merger: 

The whole world of e-commerce came to this industry before it came to other industries. 
We're growing while business models are changing on top of us. Anything we build now 
we know is wrong. You never have your feet on the ground. What internet technology 
does to the world is that you're in a continual state of aggregation and disaggregation. 
You're constantly putting stuff together and taking it apart. 

— Michael Capellas, Senior Vice President & CIO 

Implementing Enterprise Systems 

Enterprise systems implementation embodied two different challenges. One was to redefine the new 
firm's business processes. The other was to effectively manage ERP implementation projects in a large, 
complex, constantly changing organization. 

Prior to the merger, Compaq had total world-wide supply chain planning running on custom systems. 
Except in the US, this had tied into SAP for order management. IM did not immediately attempt to install 
these systems at Digital facilities, because supply chain planning would be very much affected by the 
firm's reorganization. From June to November 1998, the firm was shutting down plants, laying off 
people, and consolidating operations. Thus, prior to November 1998, there was little progress in 
integrating the supply chain across former Tandem, Digital, and Compaq sites. Ultimately, management 
had to decide the extent to which the supply chain could be standardized across geographies and product 
lines: 

In any of these large ERP projects, if you can answer 'how common is common,' you'll 
get it right. I've never seen anybody do it gracefully or painlessly. It's a very, very 
difficult question — and it's much broader than IM. It's the whole world of management 
in the 90s. Technology is allowing you to make more things common than you used to; 
but on the other hand, the flexibility of that technology is also giving capabilities at the 



14 



local level to people that never had it before. So it's a difficult balance of technology, 
management style, and just pure efficiency. 

— Michael Capellas, Senior Vice President & CIO 

In addition to supply chain differences, there were significant differences in the firms' financial closing 
and forecasting processes. For example, both Compaq and Digital used Hyperion systems for financial 
consolidation. However, Digital did the currency translation at the transaction level and then fed 
transaction data into Hyperion in U.S. dollars prior to doing cost allocations. Compaq did all of its cost 
allocations and liquidations inside its new SAP system and then sent the transactions in local currencies 
into Hyperion where they were translated. In addition, they took different approaches to foreign currency 
translation. Digital translated monthly results while Compaq translated each transaction. 

How we run our closes between Compaq and Digital Classic are very dissimilar. That 
doesn't mean one is right or one is wrong, it's just that two companies have chosen to do 
it two different ways. Basically, when you look at our systems, the only thing similar is 
they're made by the same vendor. 

— Director, Global Financial Processes 

Once the two entities moved to a common ledger system using a Compaq SAP baseline, they would have 
common allocation schemes, common reporting schemes, and common data. Even before formal approval 
of the merger, the integration team had decided to accept the Compaq SAP baseline to accomplish this 
financial and supply chain integration: 

We had really built a much higher volume, high speed environment [at Compaq]. We 
thought it would be much easier to add the complexity to the high volume than to add the 
high volume to the complexity. 

— SAP Project Manager 

The Services side of the business had some unique requirements that were not incorporated in the 
Compaq baseline. For example. Services needed the project accounting module in SAP in order to 
determine profitability and margins by work breakdown element. The integration team had intended to 
adopt Compaq Classic's SAP rapid implementation approach. However, the gradual recognition of the 
unique needs of the Services and direct sales organizations greatly increased the complexity of the 
implementation: 

It's a new world for Compaq. The old way we used to do it, when we were growing fast 
and furious, was that IM would race around and bring up SAP in big bang swoops on a 
Compaq Classic model, which was basically building PCs. Now, there's this giant 
Services arm, we have Tandem which builds giant PCs that can take a year to sell. We're 
getting too big for that now. 

— Director, Global Financial Processes 

OnePaq I A was originally intended to be simple product order integration, but it eventually grew into a 
new corporate CTO (configure-to-order) initiative. Over time, the project plan evolved to include about 
150 projects to be implemented in seven phases under OnePaq program managers in each geographic 
region. Compaq, which had traditionally defined systems projects that could be completed in three to six 
months, found itself constantly confronting scope creep: 

It got too big. We had master data issues. We also had to leam how Services was doing 
things because we had never had a Services group. We couldn't do it. Basically, we fell 
back to breaking up the project into smaller projects. In manufacturing, we're doing a 



15 



phased implementation. IM gets the nasty job of saying, 'how do I phase all of this into 
some sort of project schedule with resource constraints, cash, and all that?' 

— Director, Global Financial Processes 



Meeting the Headcount Reduction Targets 

At the same time that OnePaq implementations were getting underway, another challenge for the new IM 
organization had to be tackled; meetmg the headcount reductions filed with the SEC as part of the cost 
savings rationalization for the merger. Achieving the quotas was a non-negotiable issue, and funds had 
been put aside for termination packages. When Capellas arrived in August 1998, the 'clock was already 
ticking' but decisions about the downsizing targets for specific IM units had been delayed until his 
arrival. 

Most IT groups were given an 18% target, but a portion of the quotas would be met by attrition. Many 
experienced Digital managers — including the Digital CIO and eight other managers who had carried a 
CIO or vice president title in the Digital IS organization — had left the company before Capellas' 
appointment was announced. Additional 'cream of the crop' employees left before the layoffs began: 

As Digital had gone through rounds of cost cutting and facility closings it was not 
unusual to have one or two really good people working fi-om home in a particular area. 
Eventually this grew to a work-at-home program with hundreds of people (at least 100 in 
IS) participating. At the time of the merger, these people were viewed as a problem and 
the directive went out to either bring them into a Compaq office or lay them off. What 
many saw as progressive work practices at Digital were being eliminated. Hundreds left 
the company over this one alone. 

— Rick Fricchione, Vice President 
Advanced Information Technology and Planning 

When Capellas arrived in August, overall attrition rates were quite high. At one point 'sharks' from a 
major IT consulting firm had camped outside of the Digital offices in recruiting vans: 

You take a very hot area like IT where the job market is on fire. You compound it with an 
equally red hot market like SAP. Put it together with the uncertainty of an acquisition, 
and you will get attrition. It was also fairly clear that the new Compaq was going to be 
based in Texas, and the thought of becoming cowboys was not always that appealing to 
those long-established Bostonians. 

— Michael Capellas, Senior Vice President & CIO 

A snag in the downsizing process was caused by delays in the European union negotiations. Late in 1998, 
it became clear that it would be impossible to meet the European targets for that year. The U.S. groups 
were told to increase their 1998 layoffs. Although this meant that the U.S. targets for 1999 would be 
lowered when layoffs in Europe became possible, it impacted credibility and morale within IM: 

We ended up — after many of us had told everybody 'that was it' — having to go back and 
do another [round of layoffs]. That last one in December was pretty brutal. It had big 
morale issues. We had said we were done. 

— SAP Project Director 

The staff reductions, coupled with accelerating demands for data integration, e-coirunerce capabilities, 
and enterprise systems support, forced constant focus on attacking only the highest IM pnorities. 



16 



Transforming the Firm 

As the last year of the century began, Compaq's logo was appearing on Digital buildings in New England, 
and the re-badging of Digital Classic people was underway. IM management noted that some pieces of 
the IM integration were clear from the outset; others were evolving on a daily basis. In particular, IM's 
role was increasingly focused on enabling Compaq to transform itself from a build-to-stock to a 
configure-to-order model as the firm moved aggressively into e-business: 

We anticipated some of the ways that the new business would be put together, but some 
of the pieces were put together differently than we expected. The maturity of e-business 
and Web-based business all happened in the same time frame. The business priorities 
turned 180 degrees during the integration process. The Web project was always on the 
roadmap, but the urgency changed significantly during the first nine months of 
integration. 

— Fred Jones, Vice President, Information Services 

By early 1999, management recognized that the OnePaq initiative would not meet its initial targets, and 
management established new SAP target implementation dates for the next four quarters. These included 
merging general ledgers in Q2 of 1999; consolidating order management (in part) in Q3; integrating 
manufacturing in Q4 of 1999 and Ql of 2000; and systems integration of the Services business in Q2 of 
2000. The OnePaq initiatives were critical for addressing the goals for the merger as filed with the SEC, 
as well as for enabling Compaq to compete in an e-business world. 



Epilogue 

In July 1999 Michael Capellas, Compaq's CIO, was appointed CEO. At that fime, Compaq's revenues 
were stagnating and the firm had not yet realized the anticipated cost savings from the merger. By 
September 2000, Compaq's revenues had rebounded and operating costs were on the decline. Compaq 
reported net income of $712 million for the first two quarters of 2000 on revenues of $19.6 billion, as 
compared with net income of $97 million on $18.8 billion in revenues for the year-earlier period. In 
addition, Compaq was generating $6 million in daily sales through its website and about $50 million in 
daily sales through EDI. 

Compaq's IM unit delivered GlobalNet in mid-1999, fulfilling the green field vision established prior to 
the merger. In practice no one in the Compaq organization would be more than "two hops from the 
backbone," and the network had universal remote access. This design provided better bandwidth usage, 
which meant that Compaq could spend its dollars on the routing architecture rather than bandwidth. Bob 
Napier was brought in from GM Delphi to take the CIO role. He, in turn, placed CIOs in each of the 
business units, creating a federated IM structure for the new Compaq. 

Capellas reorganized Compaq around global business units: consumer, commercial pc's, storage, industry 
standard servers, business critical servers, professional services, and customer services. Each business 
unit appointed a vice president for key processes such as supply chain. The services unit was originally 
grouped with enterprise computing, but sales languished and Compaq lost ground to IBM. Eventually 
Capellas moved to a direct sales force rather than channel partners and he joined services with direct sales 
to better serve the needs of large corporate customers. The resulting organization was one that was 
prepared to be a total enterprise player. 



17 



Exhibit B-1 



Compaq Information Management 



Geography Support 




Solutions & Global 
Support I 



Vice Presidem 

WW Mfg. & Quality 

Systems 



Shared Services & 
Delivery i 



Fred Jones 

Vice President 

Information Services 



Michael Capellas 
Sr. Vice President & 
Chief Information OfTicer 



Director 
IM Support EMEA 



Director WW 
Sales & Marketing 



Director 
SAP Program 



Executive Assistant 



Director 
IM Support MA 



Director 
IM Support LA 



Mike Whitney 

Vice President 

WW Services. 

Solutions & Support 



Ed Pennington 

Director 

Corporate Admin 

Systems 



Director 

WW Planning 

Systems 



Rick Fricchione 

Vice President 

Advanced Information 

Technotogy & Planning 



Director 
IM Support APD 



Vice President 

Product Development 

Systems 



IM Finance 
IMHR 



Exhibit B-2 



The OnePAQ Road Map 



TANDEM ERP Program 



Integration 
of Tandem 
requirements 




Integration of Digital 
technology in 
production SAP R/3 
environments 



Integration 
of Digital 
requirements 



DIGITAL SAP Program 



OnePAQ 

integrated 
product/ 
services 
SAP R/3 
environment 



18 



1 o 



MOV « 

Date Due 



Lib-26-67 



MIT LIBRARIES 



3 9080 02246 1815